Skip to content

MemDocs v2.0.0-beta - Production-Grade Quality, 84.6% Coverage

Pre-release
Pre-release

Choose a tag to compare

View all tags
@silversurfer562 silversurfer562 released this 09 Nov 18:27
· 49 commits to main since this release
Immutable release. Only release title and notes can be modified.
v2.0.0-beta1
c097438

MemDocs v2.0.0-beta 🧠

Production-Grade Quality, Beta for User Validation

What is MemDocs?

MemDocs is a git-native memory management system that gives AI assistants persistent, project-specific memory. It generates structured, machine-readable documentation that lives in your repository—no cloud services, no recurring costs, just local/git-based storage.

📊 Why Beta? Production Quality, Early Validation

Code Quality Assessment (Third-Party Standards):

  • ✅ PyPI Production Criteria: 83% (5/6 requirements met)
  • ✅ OpenSSF Best Practices: 90% (60/67 passing badge criteria met)
  • ✅ Test Coverage: 84.6% (334 tests passing) 🎯 Exceeds 80% production threshold
  • ✅ Security: All critical vulnerabilities resolved
  • ✅ Python Support: Tested on 3.10, 3.11, 3.12

Why Beta?
We're releasing as beta not due to code quality concerns, but because we need production user validation before committing to full API stability. The codebase is production-ready—we're looking for early adopters to help validate real-world usage patterns.

Translation: The code is solid. We need you to kick the tires.

✨ Key Features

  • 🧠 Persistent Memory: AI assistants remember your project across sessions
  • 👥 Team Sharing: Memory committed to git and shared with your team
  • 💰 Zero Cost: No vector databases, no embeddings API, no subscriptions
  • ⚡ Works Offline: No cloud dependencies for retrieval
  • 🤝 Empathy Framework: Integration available via empathy package
  • 🔒 Privacy First: Optional PHI/PII detection and redaction
  • 🎯 Claude Sonnet 4.5: Latest and most powerful Claude model

🚀 Quick Start

# Install
pip install memdocs==2.0.0b0

# Set your API key
export ANTHROPIC_API_KEY="your-key-here"

# Initialize in your project
cd your-project
memdocs init

# Review a file or directory
memdocs review --path src/

# Query your project memory
memdocs query "how does authentication work?"

# View statistics
memdocs stats

📦 What's New in 2.0

Major Enhancements

  • Rebranding: DocInt → MemDocs (new name, same powerful memory)
  • Claude Sonnet 4.5: Updated to latest, most capable model
  • Modular CLI: Refactored 928-line monolith into clean command structure
  • Security Hardening: Path traversal protection, input sanitization, API key handling
  • Complete Type Hints: 100% type coverage for better IDE support
  • MCP Server: Model Context Protocol support for Claude Desktop integration

Recent Improvements

  • ✅ Wizard Exclusion: Empathy-specific code moved to empathy package (cleaner separation)
  • ✅ Improved Coverage: 84.6% (was 79% with wizards)
  • ✅ Accurate token counting with tiktoken
  • ✅ Dependency parsing for Python & JavaScript projects
  • ✅ Configurable max_tokens for Claude API
  • ✅ Python 3.10+ compatibility (tomli backport)
  • ✅ Lazy numpy imports (core features work without embeddings)

Infrastructure

  • ✅ GitHub Actions CI/CD (linting, tests, type checking)
  • ✅ Pre-commit hooks configured
  • ✅ Security policy (SECURITY.md)
  • ✅ Contributing guidelines
  • ✅ Code of Conduct

🎯 Quality Metrics

Metric Target Actual Status
Test Coverage 80% 84.6% ✅ Exceeds
Security Issues 0 0 ✅ Pass
Python Versions 3.10+ 3.10-3.12 ✅ Pass
Documentation Complete Complete ✅ Pass
CI/CD Configured Active ✅ Pass
Linting Clean 16 minor ⚠️ Acceptable

📦 Package Information

Lean & Focused: Wizards/integrations excluded from core package

  • Core MemDocs functionality only
  • Empathy integration available via: pip install empathy
  • Smaller package size, clearer separation of concerns

🐛 Known Limitations

  • CLI integration tests: Some command modules lack integration tests
  • Tree-sitter: TypeScript/JavaScript extraction planned but not yet implemented
  • Linting: 16 minor style warnings (non-blocking)

These are documented and tracked, none are blocking for basic usage.

📚 Resources

🙏 Call for Early Adopters

We're seeking early adopters to:

  • Validate the tool in real-world projects
  • Provide feedback on API ergonomics
  • Report edge cases and bugs
  • Share use cases and success stories

If you try MemDocs, please let us know how it goes! Your feedback will directly shape the 2.0.0 stable release.

📝 Full Changelog

See CHANGELOG.md for complete version history.

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Installation: pip install memdocs==2.0.0b0

Next Steps: After user validation, we'll release 2.0.0 stable with API stability guarantees.

Assets 5
Loading